Find definitions and interpretation guidance for every statistic in the Regression equation table.

Use the regression equation to describe the relationship between the response and the terms in the model. The regression equation is an algebraic representation of the regression model.

The regression equation with a binary response and more than one term takes the following form:

y' = b_{0} + b_{1}X_{1} + b_{2}X_{2} + ... + b_{k}X_{k}

In the regression equation, the letters represent the following:

- y' is the probability of the event, transformed with the link function
- b
_{0}is the constant - b
_{1}, b_{2}, ..., b_{k}are the coefficients - X
_{1}, X_{2}, ..., X_{k}are the values of the terms

Minitab displays the regression equation in uncoded units unless the model is nonhierarchical.
###### Note

For more information on hierarchy, go to What are hierarchical models?.

When the model is nonhierarchical, the regression equation is in coded units.

- Interpretation of uncoded units
- For a regression equation that is in uncoded units, interpret the coefficients using the natural units of each variable. For a categorical variable, the natural units of the variable are −1 for the low level and +1 for the high level, just as if the variable was coded. You can examine the coded coefficients in the Coefficients table. For the center point term, the variable is 1 if all of the continuous factors are at their midpoints, and is 0 otherwise. Because the equation is averaged over blocks, no coefficients for any blocks are in the equation.
- Interpretation of coded units
- For a regression equation in coded units, the low level of a factor is −1 and the high level of a factor is +1. The units for covariates are always the units in the data, even when the factors are coded. For the center point term, the variable is 1 if all of the continuous factors are at their midpoints, and is 0 otherwise. Because the equation is averaged over blocks, no coefficients for any blocks are in the equation.